The human vagina is normally colonized by billions of bacteria. In many healthy women, the dominant bacterial genus is Lactobacillus, whose members produce lactic acid and maintain a low vaginal pH.1,2 Colonization with Lactobacillus species that produce hydrogen peroxide (H2O2) has been associated with lower rates of bacterial vaginosis (BV),3 preterm birth,4 HIV acquisition,5 and higher rates of pregnancy implantation after in vitro fertilization.6 However, studies have not demonstrated the same benefit for non–H2O2-producing Lactobacillus species.
Bacterial vaginosis is a syndrome characterized both by an absence of lactobacilli and an increase in the diversity of the vaginal microbial community.7 Women with BV have an increased risk for preterm birth,8 late miscarriage,9 and HIV acquisition.5 Given the strong inverse correlation between Lactobacillus colonization and BV, it is difficult to determine whether lactobacilli are protective or whether BV is harmful. Is the presence of Lactobacillus species good or simply a marker for the absence of bad?
Many of the complications associated with BV are thought to be related to the inflammatory response to the BV-associated bacterial species.10 Higher vaginal fluid levels of interleukin (IL)-6, IL-8, and IL-1β have been associated with increased risk of HIV acquisition,11,12 as well as short cervix and preterm birth.13,14 Vaginal fluid human β-defensin (HBD2)15 and secretory leukocyte protease inhibitor (SLPI)12 levels correlate with anti-HIV activity. Bacterial vaginosis is associated with elevated IL-1β, IL-6, and IL-8, and decreased HBD2 and SLPI.16 One possible mechanism for the beneficial effect of vaginal H2O2-producing lactobacilli is that they alter the mucosal immune response to negative stimuli.17 In vitro and in models of colitis, colonization with Lactobacillus species has been associated with decreased inflammation.17,18
We hypothesized that in women with BV, the presence of vaginal H2O2-producing lactobacilli would be associated with lower levels of IL-1β, IL-6, and IL-8 and higher levels of human beta defensin 2 (HBD2) and secretory leukocyte protease inhibitor (SLPI).
MATERIALS AND METHODS
This was a secondary analysis of samples and data collected from nonpregnant women enrolled in a prospective cohort of racial disparities in preterm birth in Washington State. For the primary study, women who had delivered an early preterm infant (20–34 weeks) or a term infant (≥37 weeks) and who were US-born, King County Washington residents with no history of hypertensive complications in the preceding pregnancy were enrolled and underwent a pelvic examination. All participants signed informed consent for participation, and the study was approved by the University of Washington Institutional Review Board. Vaginal flora pattern was characterized by Gram stain using the Nugent criteria: a score of 0 to 3 indicated normal flora and 7 to 10 indicated BV. Women were tested for Neisseria gonorrhea and Chlamydia trachomatis using a combined nucleic acid amplification test (Aptima Combo 2; Gen-Probe, San Diego, CA) of vaginal fluid or urine. Trichomonas vaginalis was diagnosed by culture (In-Pouch TV; Biomed Diagnostics, White City, OR). Vaginal swabs were collected for bacterial culture and placed directly into Port-a-Cul system for transport to the laboratory, where bacteria were cultured and identified using standard techniques.19 An additional Dacron swab was saturated with vaginal fluid from the posterior fornix, placed in a sterile cryovial, eluted in 0.9 mL phosphate-buffered saline, and stored at −80°C until assayed.
For this substudy, we selected samples from women with either BV or normal vaginal microbiota by Nugent score, who also had data on the presence of H2O2-producing Lactobacillus by culture. Because of funding limitations, cultures were not performed during the whole duration of the primary study so only part of the primary cohort was eligible for this substudy. Women with intermediate flora (Nugent score 4–6) were excluded from this analysis. Four women with a positive culture for T. vaginalis were also excluded from analysis. Some of the samples selected for this substudy had already been used for previous analyses in the cohort and were not available.
Vaginal swabs were thawed, vortexed for 1 minute, and then centrifuged at 14,000×g for 10 minutes. The proinflammatory cytokines IL-6, IL-8, and IL-1ß, the mucosal defense molecule SLPI and the antimicrobial peptide HBD2 were measured in swab supernatant using standard enzyme-linked immunosorbent assay, as previously described.20,21 Values that fell below the lower limit of detection for all cytokines were assigned values of half the lower limit of detection and included in the analysis.
Eluted fluid from vaginal swabs was mixed by vortex shaker for 1 minute, and 100 μL underwent DNA extraction with the MoBio Bacteremia DNA Isolation Kit (MoBio, Carlsbad, CA). All extracted DNA was tested in a quantitative polymerase chain reaction (qPCR) using primers targeting the human 18S rRNA gene to validate that successful DNA extraction occurred. An internal amplification control PCR using exogenous DNA from a jellyfish gene was used to test for presence of PCR inhibitors. DNA was then subjected to 10 separate taxon-directed 16S rRNA gene qPCR assays for the detection and quantification of individual bacteria as previously described.22,23 Negative assays were assigned a value of half the lower limit of detection for that assay and were included in all analyses, including the calculation of mean bacterial concentrations.
Comparisons of demographic factors and immune markers between the 4 groups were made using analysis of variance, Kruskal-Wallis, or χ2. When the Kruskal-Wallis test was significant, it was followed by the Dunn test for pairwise comparisons, with Bonferroni correction for multiple comparisons. Linear regression was performed with each individual immune marker as an outcome measure, using robust standard errors and dummy variables for each of the 4 groups, with BV negative, H2O2 negative as the reference category. Additional regression analyses were performed to evaluate the effect of the quantity of Lactobacillus species by qPCR on the quantity of each immune marker. We made an a priori decision to adjust for race, as previous studies have shown differences in vaginal cytokines and vaginal microbiota between African American and white women.24,25 All analyses were performed using Stata v10.
Of 311 women enrolled in the original cohort who had a Nugent score of 0 to 3 or 7 to 10, only 92 (45%) of 203 with a normal Nugent score and 62 (59%) of 108 with BV had culture results available and were eligible for this secondary analysis. The participants with samples remaining for this analysis included 26 of 32 BV negative, H2O2-producing Lactobacillus negative (H2O2−); 47 of 60 BV-negative, H2O2+; 27 of 51 BV positive, H2O2−; and 10 of 13 BV positive, H2O2+. These 4 groups were overall demographically similar, with the exception that the BV negative, H2O2− group had a higher proportion of white participants and a lower proportion of African American participants than the other groups (Table 1). The subpopulation with available samples was not significantly different from the group of all eligible women (i.e., BV− or BV+, with culture results available; data not shown).
Detection of BV-Associated Species Differed by BV Diagnosis, But Not by Presence of H2O2-Producing Lactobacillus
Lactobacillus crispatus and Lactobacillus jensenii were detected significantly more often in BV-negative women, whereas Megasphaera, Leptotrichia/Sneathia, BVAB1, BVAB2, and BVAB3 were detected significantly more often in BV-positive women. There were no differences in detection of Lactobacillus iners, G. vaginalis, or Atopobium vaginae between the 4 groups (Table 2). When comparing BV-negative women, L. crispatus and L. jensenii were more common when culture detected H2O2-producing Lactobacillus. Among BV-positive women, only L. crispatus was significantly more common when H2O2-producing Lactobacillus were detected; L. jensenii was detected in only 2 BV-positive women, 1 in each group. Within BV diagnosis categories, there were no significant differences in detection of BV-associated species between women with and without H2O2-producing Lactobacillus by culture, although there was a trend to lower detection of Megasphaera and Leptotrichia/Sneathia when H2O2-producing lactobacilli were present.
Quantity of BV-Associated Species Varied by BV Status, But Not by Presence of H2O2-Producing Lactobacillus
In all comparisons, the quantity of bacteria was significantly different between the 4 groups, with P < 0.01. After pairwise comparisons, L. crispatus and L. jensenii were present in significantly higher quantity in the BV−, H2O2+ group compared with the other 3 groups (P ≤ 0.01 for all comparisons; Fig. 1). L. iners was present in significantly higher quantity in the BV−, H2O2− group compared with the BV−, H2O2+ group (P = 0.006), but was not significantly different from the BV+ groups. For all of the BV-associated species, the BV+ groups had significantly higher concentrations than the BV− groups, whereas comparisons within BV diagnosis category were not significantly different.
In Unadjusted Analysis, Vaginal Fluid Immune Markers Differed Between Women With and Without BV, As Well As Between Women With and Without H2O2-Producing Lactobacillus
When comparing vaginal fluid immune response markers IL-1β, SLPI and HBD2 were significantly different between the 4 groups (Fig. 2). In women with and without BV, IL-1β was lower when H2O2-producing lactobacilli were detected. The biggest difference was between the BV−, H2O2+ group compared with the BV+, H2O2− group (P = 0.003), with a trend to a significantly difference with the BV−, H2O2− group as well (P = 0.07). Women with BV had lower levels of SLPI than did women without BV whether H2O2+ (P = 0.002) or H2O2− (P = 0.02). For HBD2, the lowest value was in the BV+, H2O2− and the highest in the BV−, H2O2− group. IL-6 and IL-8 did not differ between the groups.
In Adjusted Analysis, Colonization With H2O2-Producing Lactobacillus Was Associated With Differences in IL-1β, SLPI, and HBD2
In regression analysis, adjusting for race and using BV-negative women without H2O2-producing Lactobacillus detected as a reference category, the only analyte that was significantly different for each of the other 3 groups was HBD2, which was lowest in the BV-positive women without H2O2-producing Lactobacillus detected (Table 3). SLPI was lower in BV+ women without H2O2-producing Lactobacillus detected, and IL-1β in BV−women with H2O2-producing Lactobacillus compared with the reference group.
In a regression analysis adjusted for race, higher quantities of L. crispatus were associated with significantly lower concentrations of IL-1β and higher concentrations of SLPI but no difference in HBD2 (Table 3). Higher quantities of L. jensenii were associated with higher concentrations of SLPI and HBD2. When the analysis was also adjusted for BV diagnosis, the only association that remained significant was that between quantities of L. crispatus and IL-1β. Quantity of L. iners was not associated with any of the analytes measured in either analysis.
Overall, our results suggest that H2O2-producing lactobacilli have an immunomodulatory effect in the vagina, primarily through their impact on IL-1β. In addition, it seems that this effect is independent of BV diagnosis, although BV clearly has a strong immunostimulatory effect on vaginal mucosa. Quantity of L. crispatus and L. jensenii, common H2O2 producing species, showed a correlation between bacterial quantity and inflammatory markers, whereas quantity of L. iners (a non–H2O2-producing species) showed no association with cytokine levels, suggesting that H2O2 production may be a marker for these immunomodulatory effects.
For many years, the dogma in the field suggested that Lactobacillus were beneficial because the H2O2 produced by some species killed the BV-associated bacterial species,26,27 or inactivated viruses like HIV.28 However, more recent work has demonstrated that the amount of H2O2 produced in the anaerobic environment of the vagina (as opposed to aerobic culture conditions in the laboratory) is minimal or easily neutralized29 and unlikely to be able to have the effects attributed to the beneficial Lactobacillus species.30 Some have argued that lactic acid is the true effector molecule, and O'Hanlon et al.31 showed that an acid pH induced by lactic acid, but not acetic acid, is inhibitory to BV-associated bacterial species. However, because all Lactobacillus species make lactic acid, this does not explain the clinical differences seen in epidemiologic studies between women with and without H2O2-producing species. An alternative explanation is that H2O2 production is a marker for species with some other characteristic that provides the reproductive health benefits. Our results demonstrate that this may be immunomodulation, possibly to decrease mucosal inflammation.
Data on the effects of H2O2-producing Lactobacillus species specifically on vaginal markers of mucosal immunity are limited. Anderson et al.32 measured IL-1β, IL-6, and SLPI (as well as 5 other analytes) in 47 women without BV and characterized the vaginal microbiota by culture-based methods. After adjusting for race and BMI, no differences in any cytokine markers were seen between women with and without H2O2-producing Lactobacillus species detected. In a small study of 42 Indian women with normal Nugent scores, use of a vaginal probiotic tablet containing L. brevis, L. salivarius, and L. plantarum for 8 days was associated with a decrease in vaginal IL-1β and IL-6 on day 9, which was not seen in women randomized to a placebo tablet.33 In vitro, Rose et al.17 showed that adding L. crispatus or L. jensenii to cultured vaginal epithelial cells in the presence of the toll-like receptor agonists PIC and FSL-1 decreased levels of IL-6, IL-8, and/or tumor necrosis factor α compared with the agonist alone. This effect varied by strain: a laboratory isolate of L. jensenii had less of an effect than a clinical isolate. Our results showed no impact of lactobacilli on IL-6 or IL-8, but did show a decrease in IL-1β (classically proinflammatory) and an increase in SLPI (classically anti-inflammatory), which is consistent with our hypothesis. However, the results for HBD2 demonstrate the complexity of the immune response: the highest levels of this antimicrobial peptide were seen in women who were BV negative and had no H2O2-producing lactobacilli, whereas the lowest were in women who were BV+, H2O2−. This may suggest that there is an optimal range for HBD2 and that too high or too low is undesirable.
The classic teaching in the field is that Lactobacillus species control the growth of BV-associated species.34,35 The only interventional study to examine the effect of an H2O2-producing Lactobacillus on individual vaginal BV-associated bacterial species was the phase 2A randomized trial of the H2O2-producing probiotic candidate L. crispatus CTV-05. Presence and quantity of BV-associated bacterial species were measured before and after treatment using the same assays described in our study. In women who established colonization with the probiotic, there was a significant drop in the quantity of G. vaginalis, A. vaginae, and BVAB2, which was not seen among women who did not establish colonization with the CTV-05 strain.36 Cervicovaginal fluid from healthy women has been shown to inhibit growth of Escherichia coli in vitro, and this inhibitory activity has been linked to the presence of proteins from L. crispatus and/or L. jensenii.37 Our data show no difference in quantities of BV-associated species between women with BV with and without H2O2-producing Lactobacillus species. However, women with BV and H2O2-producing Lactobacillus species detected had lower quantities of the Lactobacillus species we measured, suggesting that perhaps BV-associated species have a negative impact on the lactobacilli.
Taken together with the broader literature, our results suggest that there is an immunomodulatory effect of some H2O2-producing Lactobacillus species that is not simply due to the absence of BV-associated species. However, our results also suggest that this effect may not be robust for all markers, especially in the face of the strong inflammatory response generated by BV. In addition, our data suggest that the presence of H2O2-producing Lactobacillus alone does not have a suppressive effect on BV-associated species. Both of these associations may be more dependent on quantity of the Lactobacillus species, as suggested by the association we saw between quantity of L. crispatus and L. jensenii with vaginal immune marker concentrations. A study of the L. crispatus CTV-05 probiotic for prevention of recurrent urinary tract infections only saw a beneficial effect in women who established high-quantity vaginal colonization (>106 16S rRNA gene copies/swab).38
Our study has several limitations, including its cross-sectional design, which limits our ability to infer causal relationships. We have limited data on potential confounders that might be present such as douching or recent sexual activity. This was a secondary analysis of the remaining biologic samples from a larger cohort; this could introduce some selection bias. However, the participants with samples remaining were not significantly different than the larger cohort. We used species-specific PCR to characterize several BV-associated species, but this does not comprehensively evaluate the vaginal microbiota. Some of the differences seen may be related to species that were not tested for in this study.
Our results suggest that H2O2-producing Lactobacillus can have an immunomodulatory effect in the vagina, which is not simply due to the absence of proinflammatory bacterial species. However, the protective role of these species and their utility as probiotics may not be a simple story, as the anti-inflammatory potential is attenuated in the presence of BV-associated species and the presence of H2O2-producing Lactobacillus is not associated with lower quantities of pathogenic BV-associated species.
1. Ravel J, Gajer P, Abdo Z, et al. Vaginal microbiome of reproductive-age women. Proc Natl Acad Sci U S A 2010; 108 (suppl 1): 4680–4687.
2. O'Hanlon DE, Moench TR, Cone RA. Vaginal pH and microbicidal lactic acid when lactobacilli dominate the microbiota. PLoS One 2013; 8: e80074.
3. Vallor AC, Antonio MA, Hawes SE, et al. Factors associated with acquisition of, or persistent colonization by, vaginal lactobacilli: Role of hydrogen peroxide production. J Infect Dis 2001; 184 (11): 1431–1436.
4. Martius J, Krohn MA, Hillier SL, et al. Relationships of vaginal Lactobacillus
species, cervical Chlamydia trachomatis
, and bacterial vaginosis to preterm birth. Obstet Gynecol 1988; 71: 89–95.
5. Martin HL, Richardson BA, Nyange PM, et al. Vaginal lactobacilli, microbial flora, and risk of human immunodeficiency virus type 1 and sexually transmitted disease acquisition. J Infect Dis 1999; 180: 1863–1868.
6. Moore DE, Soules MR, Klein NA, et al. Bacteria in the transfer catheter tip influence the live-birth rate after in vitro fertilization. Fertil Steril 2000; 74: 1118–1124.
7. Fredricks DN, Fiedler TL, Marrazzo JM. Molecular identification of bacteria associated with bacterial vaginosis. N Engl J Med 2005; 353: 1899–1911.
8. Hillier SL, Nugent RP, Eschenbach DA, et al. Association between bacterial vaginosis and preterm delivery of a low-birth-weight infant. The Vaginal Infections and Prematurity Study Group. N Engl J Med 1995; 333: 1737–1742.
9. Hay PE, Lamont RF, Taylor-Robinson D, et al. Abnormal bacterial colonisation of the genital tract and subsequent preterm delivery and late miscarriage. BMJ 1994; 308: 295–298.
10. Romero R, Espinoza J, Goncalves LF, et al. The role of inflammation and infection in preterm birth. Semin Reprod Med 2007; 25: 21–39.
11. Mlisana K, Naicker N, Werner L, et al. Symptomatic vaginal discharge is a poor predictor of sexually transmitted infections and genital tract inflammation in high-risk women in South Africa. J Infect Dis 2012; 206: 6–14.
12. Morrison C, Fichorova R, Mauck C, et al. Cervical inflammation and immunity associated with hormonal contraception, pregnancy and HIV-1 seroconversion. J Acquir Immune Defic Syndr 2014; 66: 109–117.
13. Raiche E, Ouellet A, Berthiaume M, et al. Short and inflamed cervix predicts spontaneous preterm birth (COLIBRI study). J Matern Fetal Neonatal Med 2014; 27: 1015–1019.
14. Taylor BD, Holzman CB, Fichorova RN, et al. Inflammation biomarkers in vaginal fluid and preterm delivery. Hum Reprod 2013; 28: 942–952.
15. Ghosh M, Fahey JV, Shen Z, et al. Anti-HIV activity in cervical-vaginal secretions from HIV-positive and -negative women correlate with innate antimicrobial levels and IgG antibodies. PLoS One 2010; 5: e11366.
16. Mitchell C, Marrazzo J. Bacterial vaginosis and the cervicovaginal immune response. Am J Reprod Immunol 2014; 71: 555–563.
17. Rose WA 2nd, McGowin CL, Spagnuolo RA, et al. Commensal bacteria modulate innate immune responses of vaginal epithelial cell multilayer cultures. PLoS One 2012; 7: e32728.
18. Mackos AR, Eubank TD, Parry NM, et al. Probiotic Lactobacillus reuteri
attenuates the stressor-enhanced severity of Citrobacter rodentium
infection. Infect Immun 2013; 81: 3253–3263.
19. Eschenbach DA, Patton DL, Meier A, et al. Effects of oral contraceptive pill use on vaginal flora and vaginal epithelium. Contraception 2000; 62: 107–112.
20. Mitchell C, Balkus J, Agnew K, et al. Changes in the vaginal microenvironment with metronidazole treatment for bacterial vaginosis in early pregnancy. J Womens Health (Larchmt) 2009; 18: 1817–1824.
21. Mitchell C, Gottsch ML, Liu C, et al. Associations between vaginal bacteria and levels of vaginal defensins in pregnant women. Am J Obstet Gynecol 2013; 208: 132.e1–132.e7.
22. Srinivasan S, Liu C, Mitchell CM, et al. Temporal variability of human vaginal bacteria and relationship with bacterial vaginosis. PLoS One 2010; 5: e10197.
23. Fredricks DN, Fiedler TL, Thomas KK, et al. Changes in vaginal bacterial concentrations with intravaginal metronidazole therapy for bacterial vaginosis as assessed by quantitative PCR. J Clin Microbiol 2009; 47: 721–726.
24. Ryckman KK, Williams SM, Krohn MA, et al. Racial differences in cervical cytokine concentrations between pregnant women with and without bacterial vaginosis. J Reprod Immunol 2008; 78: 166–171.
25. Zhou X, Brown CJ, Abdo Z, et al. Differences in the composition of vaginal microbial communities found in healthy Caucasian and black women. ISME J 2007; 1: 121–133.
26. Patterson JL, Girerd PH, Karjane NW, et al. Effect of biofilm phenotype on resistance of Gardnerella vaginalis to hydrogen peroxide and lactic acid. Am J Obstet Gynecol 2007; 197: 170.e1–170.e7.
27. Eschenbach DA, Davick PR, Williams BL, et al. Prevalence of hydrogen peroxide-producing Lactobacillus
species in normal women and women with bacterial vaginosis. J Clin Microbiol 1989; 27: 251–256.
28. Klebanoff SJ, Kazazi F. Inactivation of human immunodeficiency virus type 1 by the amine oxidase-peroxidase system. J Clin Microbiol 1995; 33: 2054–2057.
29. O'Hanlon DE, Lanier BR, Moench TR, et al. Cervicovaginal fluid and semen block the microbicidal activity of hydrogen peroxide produced by vaginal lactobacilli. BMC Infect Dis 2010; 10: 120.
30. Muench DF, Kuch DJ, Wu H, et al. Hydrogen peroxide-producing lactobacilli inhibit gonococci in vitro but not during experimental genital tract infection. J Infect Dis 2009; 199: 1369–1378.
31. O'Hanlon DE, Moench TR, Cone RA. In vaginal fluid, bacteria associated with bacterial vaginosis can be suppressed with lactic acid but not hydrogen peroxide. BMC Infect Dis 2011; 11: 200.
32. Anderson BL, Cu-Uvin S, Raker CA, et al. Subtle perturbations of genital microflora alter mucosal immunity among low-risk pregnant women. Acta Obstet Gynecol Scand 2011; 90: 510–515.
33. Hemalatha R, Mastromarino P, Ramalaxmi BA, et al. Effectiveness of vaginal tablets containing lactobacilli versus pH tablets on vaginal health and inflammatory cytokines: A randomized, double-blind study. Eur J Clin Microbiol Infect Dis 2012; 31: 3097–3105.
34. Atassi F, Brassart D, Grob P, et al. Lactobacillus
strains isolated from the vaginal microbiota of healthy women inhibit Prevotella bivia
and Gardnerella vaginalis
in coculture and cell culture. FEMS Immunol Med Microbiol 2006; 48: 424–432.
35. Matu MN, Orinda GO, Njagi EN, et al. In vitro inhibitory activity of human vaginal lactobacilli against pathogenic bacteria associated with bacterial vaginosis in Kenyan women. Anaerobe 2010; 16: 210–215.
36. Ngugi BM, Hemmerling A, Bukusi EA, et al. Effects of bacterial vaginosis-associated bacteria and sexual intercourse on vaginal colonization with the probiotic Lactobacillus crispatus
CTV-05. Sex Transm Dis 2011; 38: 1020–1027.
37. Kalyoussef S, Nieves E, Dinerman E, et al. Lactobacillus
proteins are associated with the bactericidal activity against E. coli
of female genital tract secretions. PLoS One 2012; 7: e49506.
38. Stapleton AE, Au-Yeung M, Hooton TM, et al. Randomized, placebo-controlled phase 2 trial of a Lactobacillus crispatus
probiotic given intravaginally for prevention of recurrent urinary tract infection. Clin Infect Dis 2011; 52: 1212–1217.